Psychological Treatment of
OBSESSIVECOMPULSIVE DISORDER FUNDAMENTALS E D I T E D
AND
BEYOND
B Y
Martin M. Antony Christine Purdon Laura J. Summerfeldt
A M E R I C A N
P S Y C H O L O G I C A L A S S O C I A T I O N
W A S H I N G T O N ,
D C
Copyright © 2007 by the American Psychological Association. All rights reserved. Except as permitted under the United States Copyright Act of 1976, no part of this publication may be reproduced or distributed in any form or by any means, including, but not limited to, the process of scanning and digitization, or stored in a database or retrieval system, without the prior written permission of the publisher. Published by American Psychological Association 750 First Street, NE Washington, DC 20002 www.apa.org To order APA Order Department P.O. Box 92984 Washington, DC 20090-2984 Tel: (800) 374-2721; Direct: (202) 336-5510 Fax: (202) 336-5502; TDD/TTY: (202) 336-6123 Online: www.apa.org/books/ E-mail:
[email protected] In the U.K., Europe, Africa, and the Middle East, copies may be ordered from American Psychological Association 3 Henrietta Street Covent Garden, London WC2E 8LU England Typeset in Goudy by Stephen McDougal, Mechanicsville, MD Printer: Edwards Brothers, Inc., Ann Arbor, MI Cover Designer: Berg Design, Albany, NY Technical/Production Editor: Tiffany L. Klaff The opinions and statements published are the responsibility of the authors, and such opinions and statements do not necessarily represent the policies of the American Psychological Association. Library of Congress Cataloging-in-Publication Data Psychological treatment of obsessive-compulsive disorder : fundamentals and beyond / [edited by] Martin M. Antony, Christine Purdon, Laura Summerfeldt.—1st ed. p. cm. Includes bibliographical references and index. ISBN-13: 978-1-59147-484-5 ISBN-10: 1-59147-484-1 1. Obsessive-compulsive disorder. 2. Cognitive therapy. 1. Antony, Martin M. II. Purdon, Christine. III. Summerfeldt, Laura. [DNLM: 1. Obsessive-Compulsive Disorder—therapy. 2. Psycho-therapy—methods. 3. Obsessive-Compulsive Disorder—psychology. 4. Comorbidity. WM 176 P9737 2007] RC533.P78 2007 616.85'2270651—dc22 British Library Cataloguing-in-Publication Data A CIP record is available from the British Library. Printed in the United States of America First Edition
2006023919
CONTENTS
Contributors
ix.
Acknowledgments
xi
Introduction Martin M. Antony, Christine Purdon, and Laura J. Summerfeldt
3
I. Fundamentals of Psychological Treatment for Obsessive-Compulsive Disorder Chapter 1. Cognitive-Behavioral Models of Obsessive-Compulsive Disorder Steven Taylor, Jonathan S. Abramowitz, and Dean McKay
7 9
Chapter 2. General Issues in Psychological Treatment for Obsessive-Compulsive Disorder David F. Tolin and Gail Steketee
31
Chapter 3. Treatment Readiness, Ambivalence, and Resistance C. Alec Pollard
61
Chapter 4- Exposure and Response Prevention Karen Rowa, Martin M. Antony, and Richard P. Swinson
79
Chapter 5. Cognitive Therapy for Obsessive-Compulsive Disorder Christine Purdon II. Strategies for Specific Obsessive-Compulsive Disorder Presentations Chapter 6. Treating Contamination Concerns and Compulsive Washing David S. Riggs and Edna B. Foa Chapter 7. Treating Doubting and Checking Concerns . . . . Jonathan S. Abramowitz and Christy A. Nelson
147 149 169
Chapter 8. Treating Incompleteness, Ordering, and Arranging Concerns Laura]. Summerfeldt
187
Chapter 9. Treating Religious, Sexual, and Aggressive Obsessions S. Rachman
209
Chapter 10. Treating Compulsive Hoarding Ancy E. Cherian and Randy O. Frost III. Strategies for Specific Populations Chapter 11. Treating Obsessive-Compulsive Disorder in Children and Adolescents Martin E. Franklin, John S. March, and Abbe Garcia Chapter 12. Treating Obsessive-Compulsive Disorder in People With Poor Insight and Overvalued Ideation David Veale
vi
Ill
231
251 253
267
Chapter 13. Treating Comorbid Presentations: ObsessiveCompulsive Disorder, Anxiety Disorders, and Depression Deborah Roth Ledky, Anushka Pai, and Martin E. Franklin
281
Chapter 14- Treating Comorbid Presentations: ObsessiveCompulsive Disorder and Disorders of Impulse Control Adam S. Radomsky, Antje Bohne, and Kieron P. O'Connor
295
CONTENTS
Author Index
311
Subject Index
321
About the Editors
337
CONTENTS
vii
CONTRIBUTORS
Jonathan S. Abramowitz, PhD, Department of Psychology, University of North Carolina at Chapel Hill Martin M. Antony, PhD, ABPP, Department of Psychology, Ryerson University, Toronto, Ontario, Canada, and Anxiety Treatment and Research Centre, St. Joseph's Healthcare, Hamilton, Ontario, Canada Antje Bohne, PhD, Institute of Psychology, University of Muenster, Germany Ancy E. Cherian, PhD, Center for Anxiety and Related Disorders, Boston University, Boston, MA Edna B. Foa, PhD, Center for the Study and Treatment of Anxiety, Department of Psychiatry, University of Pennsylvania, Philadelphia Martin E. Franklin, PhD, Center for the Study and Treatment of Anxiety, Department of Psychiatry, University of Pennsylvania, Philadelphia Randy O. Frost, PhD, Department of Psychology, Smith College, Northampton, MA Abbe Garcia, PhD, Child and Adolescent Psychiatry, Brown University Medical School and Rhode Island Hospital, Providence Deborah Roth Ledley, PhD, Adult Anxiety Clinic of Temple University and private practice, Philadelphia, PA John S. March, MD, MPH, Department of Psychiatry, Duke University Medical Center, Durham, NC Dean McKay, PhD, Department of Psychology, Fordham University, Bronx, NY Christy A. Nelson, MA, Department of Psychology, University of Kansas, Lawrence
IX
Kieron P. O'Connor, PhD, Femand-Seguin Research Centre, Louis H. Lafontaine Hospital, and Department of Psychiatry, University of Montreal, Montreal, Quebec, Canada Anushka Pai, BA, Department of Psychology, University of Texas at Austin C. Alec Pollard, PhD, Anxiety Disorders Center, St. Louis Behavioral Medicine Institute, St. Louis, MO Christine Purdon, PhD, Department of Psychology, University of Waterloo, Waterloo, Ontario, Canada S. Rachman, PhD, Department of Psychology, University of British Columbia, Vancouver, British Columbia, Canada Adam S. Radomsky, PhD, Department of Psychology, Concordia University, and Centre de Recherce Fernand-Seguin, Montreal, Quebec, Canada David S. Riggs, PhD, Center for the Study and Treatment of Anxiety, Department of Psychiatry, University of Pennsylvania, Philadelphia Karen Rowa, PhD, Anxiety Treatment and Research Centre, St. Joseph's Healthcare, and Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada Gail Steketee, PhD, Boston University School of Social Work, Boston, MA Laura J. Summerfeldt, PhD, Department of Psychology, Trent University, Peterborough, Ontario, Canada Richard P. Swinson, MD, FRCP(C), FRCPsych, Anxiety Treatment and Research Centre, St. Joseph's Healthcare, and Department of Psychiatry and Behavioural Neurosciences, McMaster University, Hamilton, Ontario, Canada Steven Taylor, PhD, Department of Psychiatry, University of British Columbia, Vancouver, British Columbia, Canada David F. Tolin, PhD, Anxiety Disorders Center, Institute of Living, Hartford Hospital, Hartford, CT David Veale, MD, FRCPsych, South London and Maudsley Trust, Institute of Psychiatry, King's College London, and The Priory Hospital North London, England.
CONTRIBUTORS
ACKNOWLEDGMENTS
We thank Julia Blood for her editorial assistance on several chapters of this book. Thanks also to Susan Reynolds, Judy Nemes, and the staff at the American Psychological Association for their support. Finally, we thank the authors who contributed to this book as well as several anonymous reviewers who provided comments on a draft of the manuscript.
XI
Psychological Treatment of
OBSESSIVECOMPULSIVE DISORDER
INTRODUCTION MARTIN M. ANTONY, CHRISTINE PURDON, AND LAURA J. SUMMERFELDT
Until the mid-1960s, obsessive-compulsive disorder (OCD) was considered to be a relatively untreatable condition. In the brief span of time since then, the empirical and clinical literature have converged to point to cognitive-behavioral therapy (CBT) as among the most effective psychological interventions available for any psychiatric condition. Despite the widely recognized efficacy of behavioral and cognitive treatments for OCD, however, clinicians and researchers acknowledge that there are major challenges to implementing them successfully. In their meta-analysis of treatment outcomes for OCD, Eddy, Dutra, Bradley, and Westen (2004) observed that about half of the patients who applied for psychological treatment were excluded from the studies examined. Moreover, despite impressive group effect sizes, only about half of those who were included in treatment efficacy studies showed clinical improvement, and only one quarter recovered completely. Of equal concern, many individuals entering therapy did not complete it, although roughly two thirds of those who did complete treatment improved. In short, at each step of the treatment process, it appears that a substantial portion of the OCD population fails to benefit from the most effective psychological treatment available. One reason for this poor outcome may be
3
the fact that OCD is a complex and heterogeneous disorder that may not lend itself well to generic interventions. It is also associated with features that can substantially complicate the design, application, and delivery of treatment, such as treatment ambivalence, poor insight, and high comorbidity. Obsessive-compulsive disorder is also a disorder that occurs at many points across the life span, which again may compromise the success of generic, as opposed to tailored, treatment strategies. This volume seeks to explicate both the general, underlying features of the disorder and the unique characteristics of each subtype. The chapters cover general and specific treatment approaches, along with applications of treatment to specific populations. The book is organized into three parts. The first part covers topics that are fundamental to general cognitivebehavioral treatment of OCD. In chapter 1, Steven Taylor, Jonathan S. Abramowitz, and Dean McKay introduce cognitive-behavioral models of OCD and review empirical support for the central theoretical tenets of each. In chapter 2, David F. Tolin and Gail Steketee discuss general treatment issues in OCD, such as inpatient versus outpatient therapy, group versus individual treatment, family involvement in treatment, the role of medication in treatment, and contraindications of CBT for OCD. They provide guidelines for making important treatment decisions. In chapter 3, C. Alec Pollard discusses treatment ambivalence, readiness, and resistance in OCD, reviewing empirical data and describing a program for resolving treatment resistance and improving readiness. In chapter 4, Karen Rowa, Martin M. Antony, and Richard P. Swinson review the literature on exposure and response prevention (ERP; also known as exposure and ritual prevention) and offer a detailed, pragmatic guide for conducting ERP. This part finishes with Christine Purdon's detailed discussion in chapter 5 on implementing cognitive strategies in OCD treatment. The second part contains chapters on treating specific OCD symptom presentations. These chapters are written by leading experts in each area and feature a description of the problem, case illustrations, treatment formulations, treatment protocols, and specific troubleshooting suggestions. In chapter 6, David S. Riggs and Edna B. Foa detail treatment of contamination fears, the most common subtype in both inpatient and outpatient settings. In chapter 7, Jonathan S. Abramowitz and Christy A. Nelson describe treatment for doubting and checking concerns, which are often difficult for the average clinician to conceptualize and plan exposure for. In chapter 8, Laura J. Summerfeldt describes the treatment of incompleteness, ordering, and arranging concerns, which are less readily explained from a CBT perspective. In chapter 9, S. Rachman addresses the treatment of repugnant aggressive, sexual, and religious obsessions, which give rise to intense and diverse emotions, and yet are often concealed by patients. Finally, in chapter 10, Ancy E. Cherian and Randy O. Frost review treatment of compulsive hoarding, often considered one of the most refractory of OCD presentations. 4
ANTONY, PURDON, AND SUMMERFELDT
Chapters in the final part of this volume address issues unique to the treatment of OCD in specific populations. In chapter 11, Martin E. Franklin, John S. March, and Abbe Garcia describe applications of CBT to the treatment of OCD in children and adolescents, focusing on issues and points to consider when working with a younger population. In chapter 12, David Veale discusses issues in the conceptualization and treatment of poor insight and overvalued ideation in OCD patients. In chapter 13, Deborah Roth Ledley, Anushka Pai, and Martin E. Franklin present issues in the treatment of OCD that are comorbid with mood and anxiety problems. Finally, in chapter 14, Adam S. Radomsky, Antje Bohne, and Kieron P. O'Connor review the literature on the links between OCD and such disorders of impulse control as tics and pathological skin picking. This volume is unique in its focus on both general aspects of treatment and specific applications of CBT to diverse manifestations of OCD and to diverse populations of individuals with OCD. Each chapter is written by experts who are internationally renowned for work in their respective areas, and their treatment recommendations are based on the latest available empirical research, cognitive—behavioral theory, and extensive clinical experience. Unlike other volumes, this book acknowledges and addresses the fact that using CBT to treat OCD can be extremely challenging and full of pitfalls. This book, designed for any clinician who treats OCD, provides • a detailed, accessible exposition of the cognitive—behavioral model of treatment of OCD; • a pragmatic guide to understanding the use of behavioral and cognitive techniques in treatment of OCD; • guidelines for making important treatment decisions and improving treatment readiness; • illustrations of ways to tailor CBT techniques according to OCD subtype; and • details on issues specific to different populations of individuals with OCD. By making continued efforts to acknowledge and understand treatment obstacles and the heterogeneity of OCD, and by consistently making innovative and tailored applications of CBT to specific OCD presentations, clinicians have the chance to improve treatment efficacy dramatically. It is our hope that this volume will facilitate such improvements.
REFERENCE Eddy, K. T., Dutra, L, Bradley, R., &. Westen, D. (2004). A multidimensional metaanalysis of psychotherapy and pharmacotherapy for obsessive-compulsive disorder. Clinical Psychology Review, 24, 1011-1030. INTRODUCTION
5
I FUNDAMENTALS OF PSYCHOLOGICAL TREATMENT FOR OBSESSIVE-COMPULSIVE DISORDER
1 COGNITIVE-BEHAVIORAL MODELS OF OBSESSIVE-COMPULSIVE DISORDER STEVEN TAYLOR, JONATHAN S. ABRAMOWITZ, AND DEAN MCKAY
There are many different theories of obsessive—compulsive disorder (OCD), but comparatively few provide clear, detailed descriptions of the mechanisms thought to cause the disorder (for reviews, see D. A. Clark, 2004; Jakes, 1996; Taylor, McKay, & Abramowitz, 2005b, 2005c). Few theories have been subject to extensive empirical evaluation. Some theories account for only a subset of OCD phenomena or a subset of the empirical findings concerning the disorder. Among the most prominent theoretical approaches are the contemporary cognitive-behavioral models, which are the focus of this chapter. We begin by introducing the clinical features of OCD and illustrating the disorder with case examples. We then discuss what a good model of OCD ought to accomplish. Contemporary cognitive-behavioral models are discussed, along with a review of how well these models have performed in empirical tests of their predictions. The question of whether the models can account for all the major findings concerning OCD is also considered. We conclude with a discussion of future directions for better understanding and treating this common and often debilitating disorder.
CLINICAL FEATURES By all appearances, 34-year-old Kyle had it made; he had a great job, a loving wife, and two happy, healthy school-age kids. Yet, as he tearfully told his therapist, he was plagued with "terrible" thoughts and "stupid" habits. Whenever he came across a sharp object such as a knife or a screwdriver, he had a vivid, intrusive, and horrifying image of plunging it into the eyes of one of his children. Although the images were often triggered by the sight of sharp objects, they sometimes simply popped into his mind, seemingly out of the blue. Kyle feared that he might have some sort of unconscious desire to hurt the people he loved. To avoid triggering the upsetting thoughts, he tried to keep all the sharp objects in the house out of sight, and he often insisted that the family eat meals such as burgers or finger foods that didn't require utensils. Whenever he had one of his upsetting thoughts, which occurred on most days, he felt compelled to check four times on the safety of his children and insisted that they say the words "I'm OK, Dad." He felt deeply ashamed for continually making his children go through this ritual. For as long as she could remember, Lynda had been an anxious, overly cautious person. Now in her early 20s, she had started a new job and, for the first time, had moved out of the family home into her own rented apartment. Although she had expected to relish her newly acquired freedom, Lynda found that she was frequently preoccupied with the security of her apartment. Each day she tried to quell her many lingering doubts, which she referred to as her "sticky thinking." Every morning, when she left for work, she was beset by doubts about whether she'd "properly" locked the door. Although she checked and rechecked each morning, sometimes she set off for work only to drive back for "one more check" in an attempt to assuage her concerns. When she was at home, things were no better; she frequently was troubled by doubts about whether she'd "correctly" performed all kinds of things, such as unplugging or switching off appliances and locking windows and doors. Before going to bed each night, Lynda spent up to an hour checking that things were turned off and that her apartment was "safe and secure." Fifty-four-year-old Jim struggled with contamination problems. He was aware that other people worried about dirt and germs, but Jim knew that he was different. He felt a strong need to clean whenever he overheard foul language. Beginning in high school, he would wash until he felt that he had completely "cleaned away" thoughts of the foul word and replaced them with "good" thoughts. As he got older, the problem worsened, such that he washed even when people made disparaging remarks about people he respected (e.g., the Pope) or when people made comments about darkly powerful figures that he feared (e.g., Nostradamus). When it was inconvenient to wash immediately, Jim maintained a checklist of occurrences of unwanted thoughts, and later, when he was alone, he would wash several
10
TAYLOR, ABRAMOWTZ, AND McKAY
times to "clean away" each unwanted thought. When Jim sought treatment, his hands were red and raw from all the time spent cleaning away "bad" thoughts. These are three examples of the many faces of OCD. The disorder is characterized by obsessions or compulsions or, most typically, both (American Psychiatric Association, 2000). Obsessions are upsetting thoughts, images, or urges that intrude, unbidden, into the person's stream of consciousness. Common examples include unwanted thoughts or images of harming loved ones (as in the case of Kyle); persistent, unwarranted doubts that one has locked the door (as described by Lynda); intrusive thoughts about being contaminated (e.g., those experienced by Jim); and morally or sexually repugnant sexual thoughts (e.g., intrusive thoughts of behaving in a way that violates one's morals or runs counter to one's sexual preferences). Compulsions are repetitive, intentional behaviors or mental acts that the person feels compelled to perform, usually with a desire to resist (e.g., Jim's hand washing). Compulsions are typically intended to avert some feared event or to reduce distress. They may be performed in response to an obsession, such as repetitive hand washing evoked by obsessions about contamination. Alternatively, compulsions may be performed in accordance to certain rules, such as Kyle's checking four times that his children were unharmed. Compulsions can be overt (e.g., turning the light switch off and on 10 times) or covert (e.g., thinking a "good" thought to undo or replace a "bad" thought, as in the case of Jim). Compulsions are excessive or not realistically connected to what they are intended to prevent (American Psychiatric Association, 2000). Epidemiological surveys and factor analytic studies show that OCD is a symptomatically heterogeneous condition (McKay et al., 2004). There are four major types or constellations of OCD symptoms: (a) obsessions (aggressive, sexual, religious, or somatic) and checking compulsions; (b) symmetry obsessions and ordering, counting, and repeating compulsions; (c) contamination obsessions and cleaning compulsions; and (d) hoarding obsessions and collecting compulsions (Taylor, 2005). Obsessions and compulsions of insufficient frequency or duration to meet diagnostic criteria for OCD are common in the general population (e.g., Frost & Gross, 1993; Rachman & de Silva, 1978; Salkovskis & Harrison, 1984). Compared with clinical obsessions, those found in the general population—so-called normal obsessions—tend to be less frequent, shorter in duration, and associated with less distress (Rachman & de Silva, 1978; Salkovskis & Harrison, 1984). Normal and clinical obsessions and compulsions share common themes such as violence, contamination, and doubt (Rachman & de Silva, 1978; Salkovskis & Harrison, 1984). These similarities suggest that the study of normal obsessions and compulsions may shed light on the mechanisms of their clinically severe counterparts.
COGNITIVE-BEHAVIORAL MODELS
11
CHARACTERISTICS OF A GOOD MODEL OF OBSESSIVECOMPULSIVE DISORDER A good model of OCD should be able to do several things (Taylor et al., 2005b, 2005c). It should provide a clear description of the processes and contents of the disorder and the interactions among these conceptual elements, while being as parsimonious as reasonably possible. Thus, a good model of OCD should provide an explanation of the major clinical characteristics of the disorder (obsessions and compulsions and their interrelations) and their origins and clinical course. A good model should be able to explain the symptom heterogeneity of OCD; why do some people, for example, have checking compulsions, whereas others have contamination obsessions, and still others have hoarding rituals? A good model should be clear in its predictions. There should be no ambiguity about what counts as evidence for or against the model. The model should also lead to predictions that are falsifiable. Finally, a good model should have treatment relevance. The model should enhance our understanding of current treatments and should suggest new ways of improving treatment outcome. For example, it should be able to help clinicians understand why some treatments are effective (e.g., exposure and response prevention) and why other treatments are largely ineffective (e.g., relaxation training; Steketee, 1993). BRIEF HISTORICAL PERSPECTIVE Before the development of contemporary cognitive—behavioral models of OCD, conditioning models were the dominant explanations of the disorder, at least in the research literature. Conditioning models of OCD were based on Mowrer's (1960) two-factor model of fear (e.g., Rachman, 1971; Rachman & Hodgson, 1980; Teasdale, 1974), and they proposed that obsessional fears were acquired by classical conditioning and maintained by operant conditioning. According to these models, the obsessional fear of acquiring a serious illness from doorknobs, for example, would arise from a traumatic experience in which a loved one purportedly acquired such a disease (the unconditioned stimulus) from contact with a "dirty" doorknob in a public place (the conditioned stimulus). Obsessional fears were said to be maintained by negative reinforcement—that is, avoidance of doorknobs or compulsive washing after coming into contact with a doorknob. Here, the avoidance or compulsive ritual is negatively reinforced by the reduction in discomfort and by the reduction in the perceived probability of feared consequences such as becoming contaminated. Conditioning models led to what has been established as one of the most effective treatments for OCD: exposure and response prevention (March, Frances, Carpenter, & Kahn, 1997; also known as exposure and ritual preJ2
TAYLOR, ABRAMCWITZ, AND McKA Y
vention). This treatment involves being purposefully exposed to harmless but fear-evoking stimuli while delaying or refraining from performing the compulsive rituals. In terms of treatment implications, the conditioning models were highly fruitful; no other psychological treatment has consistently outperformed the efficacy of exposure and response prevention (Abramowitz, Taylor, & McKay, 2005). Tests of the mechanisms suggested by the model, however, were not so encouraging (Clark, 2004; Gray, 1982). Major problems include the following: • Many OCD patients do not appear to have a history of relevant conditioning experiences that might lead to obsessional fears. • The model has difficulty explaining the emergence, persistence, and content of obsessions (e.g., why would a person experience recurrent, intrusive images of strangling his or her child, even though he or she has never committed or witnessed any harm of this sort?). • OCD symptoms may change over time (e.g., a person might be compelled to check door locks and then, some weeks later, feel compelled to repetitively check on the safety of his or her spouse). • The model fails to explain why people with OCD display a broad range of levels of insight into the reasonableness of their obsessions and compulsions and why any given person's degree of insight can fluctuate across time and circumstance. These and other limitations led clinical researchers to consider cognitive explanations of OCD.
CONTEMPORARY COGNITIVE-BEHAVIORAL APPROACHES Contemporary cognitive-behavioral models of OCD fall into two broad classes: Those proposing that OCD is caused by some dysfunction in cognitive processing (general deficit models), and those postulating specific dysfunctional beliefs and appraisals as causes of obsessions and compulsions (belief and appraisal models). General Deficit Models Evidence suggests that people with OCD, compared with control participants, have deficits or abnormalities on a range of tasks, including tasks that are seemingly unrelated to threat or obsessional concerns. These findings have been shown for tasks of inductive reasoning, executive functioning (e.g., planning or set shifting), and some forms of learning and memory (Greisberg & McKay, 2003; Jurado, Junque, Vallejo, Salgado, & Grafman, COGNITIVE-BEHAVIORAL MODELS
13
2002; Woods, Vevea, Chambless, & Bayen, 2002). These deficits can persist even after successful symptomatic treatment, which suggests that the cognitive impairments are not caused by heightened anxiety or other OCD symptoms (Nielen & Den Boer, 2003). People with OCD, compared with control participants, also show weakened cognitive inhibition; that is, a weakened ability to inhibit responses, even for affectively neutral responses (e.g., Enright & Beech, 1993a, 1993b; Enright, Beech, & Claridge, 1995). Neuropsychological deficits are not found in all patients, and even when deficits are present, they tend to be mild. Nevertheless, the findings led some theorists to suggest that OCD arises from aberrations in general informationprocessing systems (e.g., Pitman, 1987; Reed, 1985) or dysfunctional reasoning processes (O'Connor, 2002). The deficits are general in the sense that they affect all information that is processed, including information related to the person's obsessional concerns (e.g., contamination stimuli) and affectively neutral information. There are five major limitations of the general deficit models. First, the models do not account for the heterogeneity of OCD symptoms (e.g., Why do some people have washing compulsions whereas others have checking rituals?). Second, the models do not account for the fact that mild neuropsychological deficits have been found in many disorders, including panic disorder, social phobia, posttraumatic stress disorder, and bulimia nervosa (Taylor, 2002); the models fail to explain why such deficits give rise to OCD instead of one of these other disorders. Third, some of the models provide only sketches of the putative mechanisms (e.g., O'Connor, 2002). Fourth, most of the models have been subject to little empirical evaluation of their predictions. Fifth, the effectiveness of exposure and response prevention in treating OCD would not be predicted from the models. If dysfunctional information processing plays any causal role in OCD, it is most likely to be a nonspecific vulnerability factor that might (or might not) play a role in obsessions and compulsions. Belief and Appraisal Models Among the most promising contemporary models of OCD are those based on Beck's (1976) cognitive specificity hypothesis, which proposes that different types of psychopathology arise from different types of dysfunctional beliefs. Major depression, for example, is said to be associated with beliefs about loss, failure, and self-denigration (e.g., "I'm a failure"). Social phobia is thought to be associated with beliefs about rejection or ridicule by others (Beck & Emery, 1985; e.g., "It's terrible to be rejected"). Panic disorder is said to be associated with beliefs about impending death, insanity, or loss of control (Beck, 1988; D. M. Clark, 1986; e.g., "My heart will stop if it beats too fast"). Several theorists have proposed that obsessions and compulsions arise from specific sorts of dysfunctional beliefs. The strength of these beliefs in14
TAYLOR, ABRAMOWITZ, AND McKA Y
fluences the person's insight into his or her OCD. Among the most sophisticated of these models is Salkovskis's cognitive-behavioral approach (e.g., Salkovskis, 1985,1989,1996) and the models based on Salkovskis's approach (e.g., Frost & Steketee, 2002). Such models form the theoretical foundations for much of the work described in later chapters of this volume. Salkovskis's model begins with the well-established finding that most people experience intrusions (i.e., thoughts, images, and impulses that intrude into consciousness) or normal obsessions. An important task for any model is to explain why almost everyone experiences cognitive intrusions (at least at some point in their lives), yet only some people experience intrusions in the form of clinical obsessions (i.e., intrusions that are unwanted, distressing, and difficult to remove from consciousness). Salkovskis (1985,1989,1996) argued that intrusions—whether wanted or unwanted—reflect the person's current concerns arising from an "idea generator" in the brain. The concerns are automatically triggered by internal or external reminders of those concerns. For example, intrusive thoughts of being contaminated may be triggered by seeing dirty objects (e.g., trash cans). Salkovskis proposed that intrusions develop into obsessions only when the individual appraises the intrusions as posing a threat for which he or she is personally responsible. An example is the intrusive image of swerving one's car into oncoming traffic. Most people experiencing such an intrusion would regard it as a meaningless cognitive event, with no harm-related implications ("mental flotsam"). Such an intrusion can develop into a clinical obsession if the person appraises it as having serious consequences for which he or she is personally responsible. The person might make an appraisal such as the following: "Having thoughts about swerving into traffic means that I'm a dangerous person who must take extra care to ensure that I don't lose control." Such appraisals evoke distress and motivate the person to try to suppress or remove the unwanted intrusion (e.g., by replacing it with a "good" thought) and to attempt to prevent any harmful events associated with the intrusion (e.g., by avoiding driving). Compulsions are conceptualized as efforts to remove intrusions and to prevent any perceived harmful consequences. Salkovskis (1985, 1989) advanced two main reasons why compulsions become persistent and excessive. First, they are reinforced by immediate distress reduction and by temporary removal of the unwanted thought (negative reinforcement, as in the conditioning models of OCD). Second, they prevent the person from learning that their appraisals are unrealistic (e.g., the person fails to learn that unwanted harm-related thoughts do not lead to acts of harm). Compulsions influence the frequency of intrusions by serving as reminders of intrusions and thereby triggering their reoccurrence. For example, compulsive hand washing can remind the person that he or she may have become contaminated. Attempts at distracting oneself from unwanted intrusions may paradoxically increase the frequency of intrusions, possibly because the distractors COGNITIVE-BEHAVIORAL MODELS
J5
become reminders (retrieval cues) of the intrusions. Compulsions can also strengthen one's perceived responsibility. That is, the absence of the feared consequence after performing the compulsion reinforces the belief that the person is responsible for removing the threat. To summarize, when a person appraises intrusions as posing a threat for which he or she is personally responsible, the person becomes distressed and attempts to remove the intrusions and prevent their perceived consequences. This reaction increases the frequency of intrusions. Thus, intrusions become persistent and distressing. In other words, they escalate into clinical obsessions. Compulsions maintain the intrusions and prevent the person from evaluating the accuracy of his or her appraisals. Why do some people, but not others, make harm- and responsibilityrelated appraisals of their intrusive thoughts? Life experiences shape the basic assumptions people hold about themselves and the world (Beck, 1976). Salkovskis (1985) proposed that assumptions about blame, responsibility, or control play an important role in OCD, as illustrated by beliefs such as "Having a bad thought about an action is the same as performing the action" and "Failing to prevent harm is the same as having caused the harm in the first place." These assumptions are thought to be acquired from a strict moral or religious upbringing or from other experiences that teach the person codes of conduct and responsibility (Salkovskis, Shafran, Rachman, & Freeston, 1999). Beyond Responsibility Although Salkovskis (e.g., 1985, 1989, 1996) emphasized the importance of responsibility appraisals and beliefs, a number of cognitive-behavioral theorists have proposed that other types of dysfunctional beliefs and appraisals are also important in OCD (e.g., Freeston, Rheaume, & Ladouceur, 1996; Frost & Steketee, 2002). Thus, contemporary cognitive-behavioral theories have extended the work of Salkovskis to propose that various types of dysfunctional beliefs and appraisals, in addition to those pertaining to responsibility, play an important role in the etiology and maintenance of OCD. Although contemporary belief and appraisal models differ from one another in some ways, their similarities generally outweigh their differences. To illustrate, Rachman (1997) proposed that "obsessions are caused by catastrophic misinterpretations of the significance of one's thoughts (images, impulses)" (p. 793). In this model, the misinterpretations are not limited to responsibility appraisals but can include any interpretation that the intrusive thought is personally significant, revealing, threatening, or even catastrophic. Such an interpretation has the effect of "transforming a commonplace nuisance into a torment" (Rachman, 1997, p. 794). The person usually interprets the intrusive thought in a personally significant way and as implying that he or she is "bad, mad, or dangerous."
16
TAYLOR, ABRAMOWITZ, AND McKAY
TABLE 1.1
Domains of Dysfunctional Beliefs Associated With Obsessive-Compulsive Disorder Belief domain Excessive responsibility Overimportance of thoughts
Need to control thoughts
Description Belief that one has the special power to cause or the duty to prevent negative outcomes. Belief that the mere presence of a thought indicates that the thought is significant. For example, the belief that the thought has ethical or moral ramifications or that thinking the thought increases the probability of the corresponding behavior or event. Belief that complete control over one's thoughts is both necessary and possible.
Overestimation of Belief that negative events are especially likely and would be threat especially awful. Perfectionism
Belief that mistakes and imperfection are intolerable.
Intolerance for uncertainty
Belief that it is necessary and possible to be completely certain that negative outcomes will not occur.
For example, a devoutly religious man experienced obscene images of Jesus with an erection on the cross whenever he tried to pray. He interpreted these images as meaning that he was "a vicious, lying hypocrite and that his religious beliefs and feelings were a sham." In another example, a man whose wife had just given birth to their first child had unwanted thoughts of beating the infant. He interpreted such thoughts as meaning that he was "dangerous and clearly unfit to be a parent." Such interpretations are thought to give rise to anxiety and dysphoria, with the consequence being intense resistance to the obsessions, attempts to suppress them, neutralization, and avoidance behavior. These examples illustrate "thought-action fusion" (Shafran, Thordarson, & Rachman, 1996), in which the person believes that his or her thoughts influence the external world (e.g., "I can cause an accident simply by thinking about one") or that thinking about a behavior is morally equivalent to performing the behavior itself (e.g., "Thinking about committing adultery is as bad as actually doing it"). Building on the work of Salkovskis, Rachman, and others, the most comprehensive contemporary belief and appraisal model is that developed by the Obsessive Compulsive Cognitions Working Group (OCCWG; Frost & Steketee, 2002). This is an international group of more than 40 investigators sharing a common interest in understanding the role of cognitive factors in OCD. The group began by developing a consensus regarding the most important beliefs (and associated appraisals) in OCD (Frost & Steketee, 2002; OCCWG, 1997). They identified responsibility beliefs and other belief domains, as listed in Table 1.1, which they conceptualized as giving rise to
COGNITIVE-BEHAVIORAL MODELS
]7
corresponding appraisals. The group developed self-report measures to assess these domains (OCCWG, 2001, 2003, 2005). In addition to the models designed to account for OCD in general, OCCWG members and others have also developed a number of "minimodels" to account for particular types of OCD symptoms, such as compulsive hoarding (Frost & Hartl, 1996; Frost, Steketee, & Williams, 2002; see also Rachman, 1997, 1998). The development of such models is consistent with the view that OCD may be etiologically heterogeneous as well as symptomatically heterogeneous (McKay et al., 2004; Taylor, 2005). The mini-models account for symptom heterogeneity in various ways, such as by proposing that particular beliefs or patterns of beliefs are important for specific types of OCD symptoms, including highly specific beliefs in addition to the broad belief domains mentioned in Table 1.1. To illustrate, compulsive hoarding is said to arise from a constellation of etiologic factors, including dysfunctional beliefs about the value of possessions (e.g., beliefs that even worthless objects might be highly valuable or useful in the future), perfectionism, intolerance of uncertainty, and difficulty making decisions (Frost & Hartl, 1996; Frost et al., 2002). These models have led to a promising new cognitive-behavioral therapy. As in exposure and response prevention, the therapy involves exposure and response prevention exercises. However, the exercises are framed as behavioral experiments to test appraisals and beliefs. To illustrate, a patient has recurrent images of terrorist hijackings and a compulsion to repeatedly telephone airports to warn them. This patient is found to hold a belief such as "Thinking about terrorist hijackings will make them actually occur." To challenge this belief, the patient and therapist can devise a test that pits this belief against a more realistic belief (e.g., "My thoughts have no influence on the occurrence of hijackings"). A behavioral experiment might involve deliberately bringing on thoughts of a hijacking and then evaluating the consequences. Cognitive restructuring methods derived from Beck's cognitive therapy (e.g., Beck & Emery, 1985) are also used to challenge OCD-related beliefs and appraisals. Empirical Tests of the Belief and Appraisal Models One of the strengths of the belief and appraisal models is that they are clearly falsifiable and have generated a large number of empirical predictions leading to a wealth of research. Twelve predictions derived from these models are listed in Table 1.2, along with a summary of their degree of empirical support. The table shows that there is encouraging support for the models, although some predictions have not been extensively evaluated and some predictions have not been supported by the research. The following sections summarize details of the findings. Space limitations preclude a detailed review of the literature; more detailed reviews can be found elsewhere (e.g., D. A. Clark, 2004; Frost & Steketee, 2002). 18
TAYLOR, ABRAMCWITZ, AND McKAY
TABLE 1.2
Twelve Predictions Derived From the Belief and Appraisal Models of Obsessive-Compulsive Disorder Empirical support
Prediction 1. The beliefs listed in Table 1.1 are distinct from one another. 2. The beliefs statistically predict or are correlated with OCD symptoms. 3. The beliefs should show specificity; they should be more strongly correlated with OCD symptoms than with measures of general distress (i.e., depression and general anxiety). 4. The beliefs interact with one another to statistically predict OCD symptoms. 5. OCD patients should generally score higher than control participants on measures of beliefs and appraisals. 6. Experimental manipulations of appraisals (e.g., increases or decreases in responsibility appraisals) lead to corresponding changes in OCD symptoms. 7. Naturally occurring events that increase the strength of beliefs or occurrence of appraisals (e.g., events increasing perceived responsibility) lead to increases in OCD symptoms. 8. OCD patients report learning histories that could give rise to the beliefs. 9. Efforts to suppress unwanted intrusive thoughts lead to an increased frequency of these thoughts. 10. Treatment-related reductions in OCD symptoms are associated with reductions in the strength of beliefs and frequency of appraisals. 11. Treatments that directly target beliefs and appraisals (e.g., cognitive-behavior therapy) are more effective than treatments that do not directly target these factors (e.g., exposure and response prevention). 12. Treatments that directly target beliefs and appraisals should be more tolerable for OCD patients (i.e., there should be fewer treatment dropouts).
++ +
++ + +
? + + -
+
Note. + = Preliminary support; ++ = strong support; - = not supported; ? = not yet adequately tested. Beliefs and appraisals refer to OCD-related beliefs and appraisals, such as those listed in Table 1.1.
Prediction 1: Beliefs should be distinguishable from one another. The first prediction states that the beliefs listed in Table 1.1 should be empirically distinguishable from one another. If beliefs about inflated responsibility, for example, play a specific role in OCD, then it should be possible to demonstrate that the effects of responsibility can be empirically disentangled from other beliefs. In other words, the beliefs should not be so highly correlated with one another that they form a single nonspecific or general OCD belief factor. The research does not support this prediction. Some research has examined the factor structure of two measures: the Obsessive Beliefs Questionnaire (OBQ; a measure of each belief domain listed in Table 1.1), and the Interpretation of Intrusions Inventory (III; a measure of appraisals of intrusive thoughts in COGNITIVE-BEHAVIORAL MODELS
19
which three appraisal domains are assessed—responsibility, importance of thoughts, and control of thoughts). Factor analytic research of the III indicates that the scale is unifactorial instead of consisting of the three predicted factors (OCCWG, 2005). Factor analyses of the OBQ indicate that it consists of three factors—inflated personal responsibility and the tendency to overestimate threat, perfectionism and intolerance of uncertainty, and overimportance and overcontrol of thoughts—instead of the predicted six (OCCWG, 2005). Hierarchical factor analysis of the OBQ indicates that these factors load on a single high-order factor and that the three factors account for a small proportion of the variance in OBQ scores (6%-7%) once the higher-order factor is taken into consideration (Taylor, McKay, & Abramowitz, 2005a). Thus, the findings raise the question about the merits of distinguishing among the various belief and appraisal domains such as those listed in Table 1.1. Predictions 2 and 3: Beliefs should predict OCD symptoms and show sped' ficity in correlations. The second and third predictions have received more support, both for the individual scales of the OBQ and III and for their factor scores. The scales and factors are each correlated with measures of OCD symptoms. The correlations with OCD symptoms tend to be larger than correlations with measures of general distress (i.e., depression and general anxiety), and the correlations with OCD symptoms remained significant even when the effects of general distress were partialled out (OCCWG, 2001, 2003). Several other studies using the OBQ or similar measures have also shown that these sorts of dysfunctional beliefs are correlated with many forms of OCD symptoms (e.g., Foa, Sacks, Tolin, Przeworski, & Amir, 2002; Tolin, Abramowitz, Brigidi, & Foa, 2003; Tolin, Woods, & Abramowitz, 2003). Prediction 4: Beliefs should interact to predict OCD symptoms. Belief and appraisal models predict that beliefs (and possibly appraisals) should interact with one another to give rise to obsessions and compulsions; as D. A. Clark (2004) observed, Dysfunctional beliefs and appraisals involved in the pathogenesis of obsessions are complex.... Simply defining the cognitive basis of OCD in terms of single constructs will obfuscate the true, complex, interactive and multidimensional nature of cognition in OCD. (p. 109)
To illustrate the potential interactions of beliefs, one's sense of personal responsibility could influence the perceived importance of controlling one's thoughts so that harm does not occur. Alternatively, beliefs about the importance of one's thoughts (T) might inflate responsibility (R) beliefs (Thordarson & Shafran, 2002). If one conceptualizes this in tetms of a path diagram, T and R could have direct effects on OCD symptoms, and T would also have an indirect (interactive) effect via its influence on R. Perfectionism (P) and T might also interact. According to Salkovskis et al. (2000), "Perfectionism is usually defined in terms which suggest more enduring personality-type characteristics, which might be expected to interact with the 20
TAYLOR, ABRAMOWITZ, AND McKAY
appraisal of intrusions, particularly when such intrusions concern the completion (or non completion) of particular actions" (p. 364). Responsibility might also inflate P (Salkovskis & Forrester, 2002). To test these predictions, we conducted a series of regression analyses in which the main effects for each belief (R, P, T) and their two- and three-way interactions were entered as predictors of measures of OCD symptoms. Main effects were significant predictors, but the interactions were not (Taylor, Abramowitz, & McKay, 2005). The findings suggest that the cognitive-behavioral models can be simplified to include only main effects. Prediction 5: OCD patients should score higher than control participants. Research generally supports the prediction that OCD patients, compared with clinical and nonclinical control participants, score highest on the OBQ and III (OCCWG, 2003, 2005), although some of these results are trends (p < .10; Taylor et al, 2005a). Prediction 6: Experimental manipulations of appraisals should influence OCD symptoms. A handful of studies have experimentally manipulated OCD-related appraisals, particularly responsibility appraisals, to assess the effects on compulsive checking (e.g., Bouchard, Rheaume, &. Ladouceur, 1999; Lopatka &Rachman, 1995; Rachman, Shafran, Mitchell, Trant, &Teachman, 1996). Research suggests that checking is more frequent when high responsibility is induced (e.g., for checking that a stove is turned off), compared with when low responsibility is induced. Prediction 7: Events that strengthen beliefs or appraisals should increase OCD symptoms. A small number of studies have examined whether naturally occurring events that influence OCD-related beliefs or appraisals are related to the development or exacerbation of obsessions and compulsions. Childbirth, for example, increases the sense of personal responsibility for both parents. The increase in responsibility has been associated with the onset or exacerbation of OCD symptoms, at least in some individuals (Abramowitz, Khandker, Nelson, Deacon, & Rygwall, 2005; Abramowitz, Moore, Carmin, Wiegartz, & Purdon, 2001). Prediction 8: Particular learning histories contribute to the development of OCD-related dysfunctional beliefs. Belief and appraisal models emphasize the role of learning experiences purported to give rise to the development of dysfunctional beliefs ("mal-learning") such as those listed in Table 1.1. This suggests that it should be possible to identify such learning experiences in people with OCD. This prediction has not been systematically investigated, although case studies have described such learning experiences (e.g., de Silva & Marks, 2001; Salkovskis et al., 1999; Tallis, 1994). Examples include a childhood environment that encouraged the development of rigid or extreme codes of conduct (thereby giving rise to inflated responsibility) and events in which one's thoughts were correlated with a serious misfortune (e.g., wishing that someone would die and then learning that the person had died from some mishap), which could lead to the development of beliefs about the importance COGNITIVE-BEHAVIORAL MODELS
21
of controlling one's thoughts. Controlled research is needed to determine whether most people with OCD report such experiences and whether they are more likely to have these experiences than control participants. Prediction 9: Excessive attempts to control OCD symptoms should worsen these systems. Belief and appraisal models propose that OCD is maintained, in part, by trying too hard to control one's unwanted thoughts or by trying too hard to allay one's doubts. Consistent with this, experimental evidence suggests that repetitive checking actually increases doubt and uncertainty (van den Hout & Kindt, 2003a, 2003b). The research on attempts to control unwanted thoughts has yielded a more complex pattern of results. Experimental studies of non-OCD participants suggest that deliberate attempts to suppress unwanted thoughts often (but not invariably) lead to a paradoxical increase in the frequency of these thoughts (Wenzlaff & Wegner, 2000). Given the degree to which people with OCD strive to avoid their unwanted thoughts, this suggests that deliberate attempts to suppress obsessions should paradoxically increase the frequency of obsessions. There is inconsistent evidence that this occurs in OCD, although research indicates that people with OCD symptoms are more likely to try to suppress their unwanted, intrusive thoughts (Purdon, 2004). Predictions 10, II, and 12: Targeting beliefs and appraisals should improve treatment outcome. The final set of predictions concern the treatment relevance of the belief and appraisal models of OCD. Belief and appraisal models underscore the importance of cognitive factors in maintaining OCD and also predict that interventions that reduce the strength of OCD-related dysfunctional beliefs (e.g., the overestimation of threat) should improve treatment outcome. Reducing the strength of these beliefs should also lead patients to be more willing to engage in behavioral and cognitive—behavioral treatments that encourage them to confront the things that they fear, such as exposure to contaminants or to refrain from performing rituals that they believe will avert feared consequences (e.g., by refraining from compulsively repeating a prayer after having a "bad" thought about a family member). Consistent with the belief and appraisal models, studies have shown that treatments that reduce OCD symptoms also reduce the strength of OCDrelated beliefs (Bouvard, 2002; Emmelkamp, van Oppen, & van Balkom, 2002; McLean et al, 2001). Treatments that directly target OCD-related beliefs (i.e., cognitive—behavioral therapy) are associated with a lower proportion of dropouts than treatments that do not directly target these beliefs, such as exposure and response prevention (Abramowitz, Taylor, & McKay, 2005). However, cognitive-behavior therapy for OCD is no more effective than exposure and response prevention (Abramowitz, Taylor, & McKay, 2005). The latter finding might challenge the belief and appraisal models of OCD. Alternatively, these findings may simply indicate that cognitivebehavioral therapy is not as powerful a vehicle of belief changes as is expo-
22
TAYLOR, ABRAMOWITZ, AND McKAY
sure and response prevention. As Bandura (1977) mentioned many years ago, behavioral interventions (e.g., exposure and response prevention) may be the most potent agents of cognitive change. Overall, the predictions summarized in Table 1.2 have mixed but generally positive support. Even so, we agree with D. A. Clark's (2004) conclusion that more research needs to be done to firmly establish that beliefs and appraisals play a causal role in OCD. Neglected Realms of Research As we described in the previous section, many of the predictions derived from belief and appraisal models have received encouraging empirical support. The predictions that were not supported suggest possible avenues for refining the models. As such, the models can be regarded as open concepts (Meehl, 1977), which are amenable to development and change in response to empirical findings. A limitation of the belief and appraisal models is that they largely ignore the burgeoning research literature on the neuropsychology and neurobiology of OCD. It is unclear, for example, how the various neuropsychological deficits and reasoning abnormalities are related, if at all, to dysfunctional beliefs and appraisals in OCD. An exception is Frost's mini-model of hoarding (Frost & Hartl, 1996; Frost et al, 2002), which describes how informationprocessing abnormalities, such as decision-making difficulties, might be related to dysfunctional beliefs and appraisals. Yet even this model neglects the extensive research on the neurobiology of OCD. The brain obviously forms the organic foundation from which beliefs, appraisals, and "idea generators" emerge. And brain structures and circuits are influenced by genetic factors. Thus, a more complete understanding of the etiology of OCD may arise if the belief and appraisal models can be integrated with neurobiological and genetic research. Important questions include the following: How can belief and appraisal models be reconciled with neuroimaging research, which shows that OCD is associated with structural aberrations (e.g., volumetric abnormalities) and functional brain abnormalities, including abnormalities in the orbital frontal cortex and basal ganglia (e.g., Pujol et al., 2004; Szeszko et al., 1999; Whiteside, Port, & Abramowitz, 2004)? How can the belief and appraisal models be reconciled with research suggesting that OCD sometimes abruptly emerges in previously normal people after streptococcal infection and abates when the infection is treated (e.g., Swedo, 2002)? Belief and appraisal models emphasize the importance of various forms of mal-learning in the development of dysfunctional beliefs and appraisals in OCD. Yet what about the role of genetic factors? Research shows that various forms of beliefs are heritable, including religious and political ideologies
COGNITIVE-BEHAVIORAL MODELS
23
(e.g., Rowe, 1994), and dysfunctional beliefs implicated in various forms of psychopathology have also been shown to be heritable (e.g., Jang, Stein, Taylor, & Livesley, 1999; Taylor, Thordarson, Jang, & Asmundson, 2006). Thus, the question arises as to the relative importance of genetic and environmental factors in OCD-related beliefs and appraisals.
CONCLUSION The belief and appraisal models of OCD have many of the properties that a good model ought to have; for example, the models are falsifiable, make clear predictions, and have treatment relevance. Not surprisingly, the models have led to a rich program of research into the etiology and treatment of OCD. Although there is a good deal of empirical support for belief and appraisal conceptualizations, these models have also encountered some difficulties, such as failures to empirically support some predictions. These models are works in progress, and no doubt they will be refined in the coming years to deal with these obstacles. A more important concern, however, is that the models have been developed largely in a cognitive-behavioral vacuum; that is, they have ignored the mounting body of research on the importance of neurobiological and genetic factors in OCD. A more complete understanding of this disorder is likely to arise if theorists and researchers are willing to tackle the challenging task of integrating mind and brain— that is, beliefs and appraisals with neuroscience. Such efforts may eventually lead to a comprehensive model of OCD. Another potentially important avenue of research is to extend the conceptual and empirical work on OCD subtypes. It is possible that the belief and appraisal models apply only to some forms of OCD. Indeed, some research suggests that some OCD patients have essentially normal scores on dysfunctional beliefs listed in Table 1.1 (Taylor, Abramowitz, McKay, Calamari, et al., 2005). Some models of OCD do not regard dysfunctional beliefs as playing an important role (Jakes, 1996; Swedo, 2002; Szechtman &. Woody, 2004). Swede's (2002) model, for example, proposes that some cases of OCD, as well as some other disorders, arise from pediatric streptococcal infection that damages the basal ganglia and associated structures. Szechtman and Woody (2004) suggested that OCD arises from a dysfunction in a noncognitive and emotion-based security motivation system located in the brain. Neither of these models includes dysfunctional beliefs as explanatory constructs. It is possible that different theoretical models apply to different subtypes of OCD. That is, models emphasizing the role of dysfunctional beliefs and appraisals might apply only to a subgroup of cases of OCD or to particular symptom presentations. Further research is needed to explore this intriguing possibility. 24
TAYLOR, ABRAMOWITZ, AND McKAY
REFERENCES Abramowitz, J. S., Khandker, M.,Nelson, C. A., Deacon, B. ]., &Rygwall, R. (2005). The role of cognitive factors in the pathogenesis of obsessions and compulsions: A prospective study. Manuscript submitted for publication. Abramowitz, J. S., Moore, K., Carmin, C., Wiegartz, P., & Purdon, C. (2001). Acute onset of obsessive-compulsive disorder in males following childbirth. Psychosomatics, 42, 429-431. Abramowitz, J. S., Taylor, S., & McKay, D. (2005). Potentials and limitations of cognitive therapy for obsessive-compulsive disorder. Cognitive Behaviour Therapy, 34, 140-147. American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). Washington, DC: Author. Bandura, A. (1977). Self-efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84, 191-215. Beck, A. T. (1976). Cognitive therapy and the emotional disorders. New York: International Universities Press. Beck, A. T. (1988). Cognitive approaches to panic disorder: Theory and therapy. In S. Rachman &.J. D. Maser (Eds.), Panic: Psychological perspectives (pp. 91-109). Hillsdale, NJ: Erlbaum. Beck, A. T., & Emery, G. (1985). Anxiety disorders andphobias: A cognitive perspective. New York: Basic Books. Bouchard, C., Rheaume, J., & Ladouceur, R. (1999). Responsibility and perfectionism in OCD: An experimental study. Behaviour Research and Therapy, 37, 239248. Bouvard, M. (2002). Cognitive effects of cognitive-behavior therapy for obsessive compulsive disorder. In R. O. Frost & G. S. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment, and treatment (pp. 403-416). Oxford, England: Elsevier. Clark, D. A. (2004). Cognitive-behavioral therapy for OCD. New York: Guilford Press. Clark, D. M. (1986). A cognitive approach to panic. Behaviour Research andTherapy, 24, 461-470. de Silva, P., & Marks, M. (2001). Traumatic experiences, post-traumatic stress disorder and obsessive-compulsive disorder. International Review of Psychiatry, 13, 172-180. Emmelkamp, P. M. G., van Oppen, P., & van Balkom, A. J. (2002). Cognitive changes in patients with obsessive compulsive rituals treated with exposure in vivo and response prevention. In R. O. Frost & G. S. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment, and treatment (pp. 391-401). Oxford, England: Elsevier. Enright, S. J., &. Beech, A. R. (1993a). Further evidence of reduced inhibition in obsessive compulsive disorder. Personality and Individual Differences, 14, 387395. COGNITIVE-BEHAVIORAL MODELS
25
Enright, S. J., & Beech, A. R. (1993b). Reduced cognitive inhibition in obsessive compulsive disorder. British Journal of Clinical Psychology, 32, 67-74Enright, S.J., Beech, A. R., &Claridge, G. S. (1995). The locus of negative priming effects in obsessive-compulsive disorder and other anxiety disorders. Personality and Individual Differences, 19, 535-542. Foa, E. B, Sacks, M. B., Tolin, D. R, Przeworski, A., & Amir, N. (2002). Inflated perception of responsibility for harm in OCD patients with and without checking compulsions: A replication and extension. Journal of Anxiety Disorders, 16, 443-453. Freeston, M. H., Rheaume, J., & Ladouceur, R. (1996). Correcting faulty appraisals of obsessional thoughts. Behaviour Research and Therapy, 34, 433-446. Frost, R. O., & Gross, R. C. (1993). The hoarding of possessions. Behaviour Research andTherapy, 31,367-381. Frost, R. O., & Hartl, T. L. (1996). A cognitive-behavioral model of compulsive hoarding. Behaviour Research and Therapy, 34, 341-350. Frost, R. O., & Steketee, G. (2002). Cognitive approaches to obsessions and compulsions: Theory, assessment and treatment. Oxford, England: Elsevier. Frost, R. O., Steketee, G., & Williams, L. (2002). Compulsive buying, compulsive hoarding, and obsessive-compulsive disorder. Behavior Therapy, 33, 201-214Gray, J. A. (1982). The neuropsychology of anxiety. London: Oxford University Press. Greisberg, S., & McKay, D. (2003). Neuropsychology of obsessive-compulsive disorder: A review and treatment implications. Clinical Psychology Review, 23, 95117. Jakes, I. (1996). Theoretical approaches to obsessive-compulsive disorder. Cambridge, England: Cambridge University Press. Jang, K. L., Stein, M. B., Taylor, S., & Livesley, W. J. (1999). Gender differences in the etiology of anxiety sensitivity: A twin study. Journal of Gender Specific Medicine, 2, 39-44Jurado, M. A., Junque, C., Vallejo, J., Salgado, P., & Grafman, J. (2002). Obsessivecompulsive disorder (OCD) patients are impaired in remembering temporal order and in judging their own performance. Journal of Clinical and Experimental Neuropsychology, 24, 261-269. Lopatka, C., &Rachman, S. (1995). Perceived responsibility and compulsive checking: An experimental analysis. Behaviour Research and Therapy, 33, 674—684March, J. S., Frances, A., Carpenter, L. L., & Kahn, D. (1997). Expert consensus treatment guidelines for obsessive-compulsive disorder: A guide for patients and families. Journal of Clinical Psychiatry, 58(Suppl. 4), 65-72. McKay, D., Abramowitz, J. S., Calamari, J., Kyrios, M., Sookman, D., Taylor, S., & Wilhelm, S. (2004). A critical evaluation of obsessive-compulsive disorder subtypes: Symptoms versus mechanisms. Clinical Psychology Review, 24, 283-313, McLean, P. D., Whittal, M. L., Thordarson, D., Taylor, S., Sochting, I., Koch, W. J., et al. (2001). Cognitive versus behavior therapy in the group treatment of obsessive compulsive disorder. Journal of Consulting and Clinical Psychology, 69, 205-214. 26
TAYLOR, ABRAMOWITZ, AND McKAY
Meehl, P. E. (1977). Specific etiology and other forms of strong influence: Some quantitative meanings. Journal of Medicine and Philosophy, 2, 33-53. Mowrer, O. H. (1960). Learning theory and behavior. New York: Wiley. Nielen, M. M. A., &. Den Boer, J. A. (2003). Neuropsychological performance of OCD patients before and after treatment with fluoxetine: Evidence for persistent cognitive deficits. Psychological Medicine, 33, 917-925. Obsessive Compulsive Cognitions Working Group. (1997). Cognitive assessment of obsessive-compulsive disorder. Behaviour Research and Therapy, 35, 667-681. Obsessive Compulsive Cognitions Working Group. (2001). Development and initial validation of the Obsessive Beliefs Questionnaire and the Interpretation of Intrusions Inventory. Behaviour Research and Therapy, 39, 987-1005. Obsessive Compulsive Cognitions Working Group. (2003). Psychometric validation of the Obsessive Beliefs Questionnaire and the Interpretation of Intrusions Inventory: Part 1. Behaviour Research and Therapy, 41, 863-878. Obsessive Compulsive Cognitions Working Group. (2005). Psychometric validation of the Obsessive Beliefs Questionnaire and the Interpretation of Intrusions Inventory: Part 2. Behaviour Research and Therapy, 43, 1527-1542. O'Connor, K. (2002). Intrusions and inferences in obsessive compulsive disorder. Clinical Psychology and Psychotherapy, 9, 38-46. Pitman, R. A. (1987). A cybernetic model of obsessive-compulsive pathology. Comprehensive Psychiatry, 28, 334-343. Pujol, J., Soriano-Mas, C., Alonso, P., Cardoner, N., Menchon, J. M., Deus, J., &. Vallejo, J. (2004). Mapping structural brain alterations in obsessive-compulsive disorder. Archives of General Psychiatry, 61, 720-730. Purdon, C. (2004). Empirical investigations of thought suppression in OCD. Journal of Behavior Therapy and Experimental Psychiatry, 35, 121-136. Rachman, S. (1971). Obsessional ruminations. Behaviour Research and Therapy, 9, 225-238. Rachman, S. (1997). A cognitive theory of obsessions. Behaviour Research and Therapy, 35, 793-802. Rachman, S., & de Silva, P. (1978). Abnormal and normal obsessions. Behaviour Research and Therapy, 16, 233-248. Rachman, S., Si Hodgson, R. J. (1980). Obsessions and compulsions. Englewood Cliffs, NJ: Prentice Hall. Rachman, S., Shafran, R., Mitchell, D., Trant, J., & Teachman, B. (1996). How to remain neutral: An experimental analysis of neutralization. Behaviour Research and Therapy, 34, 889-898. Reed, G. F. (1985). Obsessional experience and compulsive behaviour: A cognitive structural approach. New York: Academic Press. Rowe, D. C. (1994). The limits of family influence: Genes, experience, and behavior. New York: Guilford Press. Salkovskis, P. M. (1985). Obsessional—compulsive problems: A cognitive—behavioural analysis. Behaviour Research and Therapy, 25, 571-583. COGNITIVE-BEHAVIORAL MODELS
27
Salkovskis, P. M. (1989). Cognitive-behavioural factors and the persistence of intrusive thoughts in obsessional problems. Behaviour Research and Therapy, 27, 677-682. Salkovskis, P. M. (1996). Cognitive-behavioral approaches to the understanding of obsessional problems. In R. M. Rapee (Ed.), Current controversies in the anxiety disorders (pp. 103-134). New York: Guilford Press. Salkovskis, P. M., & Forrester, E. (2002). Responsibility. In R. O. Frost & G. S. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment and treatment (pp. 45-61). Oxford, England: Elsevier. Salkovskis, P. M., & Harrison, J. (1984). Abnormal and normal obsessions: A replication. Behaviour Research and Therapy, 22, 1-4. Salkovskis, P. M., Shafran, R., Rachman, S., & Freeston, M. H. (1999). Multiple pathways to inflated responsibility in obsessional problems: Possible origins and implications for therapy and research. Behaviour Research and Therapy ,37, 10551072. Salkovskis, P. M., Wroe, A. L, Gledhill, A., Morrison, N., Forrester, E., Richards, C., et al. (2000). Responsibility attitudes and interpretations are characteristic of obsessive compulsive disorder. Behaviour Research and Therapy, 38, 347-372. Shafran, R., Thordarson, D. S., & Rachman, S. (1996). Thought-action fusion in obsessive-compulsive disorder. Journal of Anxiety Disorders, 10, 379-391. Steketee, G. S. (1993). Treatment of obsessive compulsive disorder. New York: Guilford Press. Swedo, S. E. (2002). Pediatric autoimmune neuropsychiatric disorders associated with streptococcal infections (PANDAS). Molecular Psychiatry, 7, S24-S25. Szechtman, H., & Woody, E. (2004). Obsessive-compulsive disorder as a disturbance of security motivation. Psychological Review, 111, 111-127. Szeszko, P. R., Robinson, D., Alvir, ]. M., Bilder, R. M., Lencz, T., Ashtari, M., et al. (1999). Orbital frontal and amygdala volume reductions in obsessive-compulsive disorder. Archives of General Psychiatry, 56, 913-919. Tallis, F. (1994). Obsessions, responsibility and guilt: Two case reports suggesting a common and specific aetiology. Behaviour Research and Therapy, 32, 143—145. Taylor, S. (2002). Cognition in obsessive-compulsive disorder: An overview. In R. O. Frost & G. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment, and treatment (pp. 1-12). Oxford, England: Elsevier. Taylor, S. (2005). Dimensional and categorical models of OCD: A critical analysis. In J. S. Abramowitz & A. C. Houts (Eds.), Handbook of OCD: Concepts and controversies (pp. 21—41). New York: Kluwer Academic. Taylor, S., Abramowitz, J. S., & McKay, D. (2005). Are there interactions among dysfunctional beliefs in obsessive compulsive disorder? Cognitive Behaviour Therapy, 34, 89-98. Taylor, S., Abramowitz, ]. S., McKay, D., Calamari, J. E., Sookman, D., Kyrios, M., et al. (2005). Do dysfunctional beliefs play a role in all types of obsessive-compulsive disorder? Journal of Anxiety Disorders, 20, 85-97.
28
TAYLOR, ABRAMOWITZ, AND McKAY
Taylor, S., McKay, D., & Abramowitz, J. S. (2005a). Hierarchical structure of dysfunctional beliefs in obsessive-compulsive disorder. Cognitive Behaviour Therapy, 34, 216-228. Taylor, S., McKay, D., & Abramowitz, J. S. (2005b). Is obsessive-compulsive disorder a disturbance of security motivation? Comment on Szechtman and Woody (2004). Psychological Review, 112, 656-657. Taylor, S., McKay, D., & Abramowitz, J. S. (2005c). Problems with the security motivation model remain largely unresolved: Response to Woody and Szechtman (2005). Psychological Review, 112, 656-657. Taylor, S., Thordarson, D. S., Jang, K. L, & Asmundson, G. J. G. (2006). Genetic and environmental origins of health anxiety: A twin study. World Psychiatry, 5, 47-50. Teasdale, J. D. (1974). Learning models of obsessional-compulsive disorder. In H. R. Beech (Ed.), Obsessional states (pp. 197-229). London: Methuen. Thordarson, D. S., & Shafran, R. (2002). Importance of thoughts. In R. O. Frost & G. S. Steketee (Eds.), Cognitive approaches to obsessions and compulsions: Theory, assessment and treatment (pp. 15-28). Oxford, England: Elsevier. Tolin, D. F., Abramowitz, J. S., Brigidi, B. D., & Foa, E. B. (2003). Intolerance of uncertainty in obsessive-compulsive disorder. Journal of Anxiety Disorders, 17, 233-242. Tolin, D. F., Woods, C. M., & Abramowitz, J. S. (2003). Relationship between obsessive beliefs and obsessive-compulsive symptoms. Cognitive Therapy and Research, 27, 657-669. van den Hout, M., & Kindt, M. (2003a). Phenomenological validity of an OCDmemory model and the remember/know distinction. Behaviour Research and Therapy, 41, 369-37'8. van den Hout, M., & Kindt, M. (2003b). Repeated checking causes memory distrust. Behaviour Research and Therapy, 41, 301-316. Wenzlaff, R. M., & Wegner, D. M. (2000). Thought suppression. Annual Review of Psychology, 51, 59-91. Whiteside, S. P., Port, J. D., & Abramowitz, J. S. (2004). A meta-analysis of functional neuroimaging in obsessive-compulsive disorder. Psychiatry Research: Neuroimaging, 132, 69-79. Woods, C. M., Vevea, J. L., Chambless, D. L., &. Bayen, U. J. (2002). Are compulsive checkers impaired in memory? A meta-analytic review. Clinical Ps^chol Science and Practice, 9, 353-366.
COGNITIVE-BEHAVIORAL MODELS
29
2 GENERAL ISSUES IN PSYCHOLOGICAL TREATMENT FOR OBSESSIVECOMPULSIVE DISORDER DAVID F. TOLIN AND GAIL STEKETEE
The most widely tested psychological treatment for obsessivecompulsive disorder (OCD) is exposure and response prevention (ERP; also known as exposure and ritual prevention), in which patients are gradually exposed to their obsessive feared situations and asked not to engage in rituals or avoidance behaviors. This intervention typically leads to reductions in negative emotions such as anxiety, guilt, and depression; to shifts in beliefs about the probability of harm; and to increased tolerance of intrusive thoughts. It also reduces avoidance behaviors and mental and behavioral rituals that reinforce the obsessive thoughts. Formal cognitive therapy (based on Beck's model; Beck, Emery, & Greenberg, 1985) may also provide relief from OCD symptoms, as do serotonergic medications. This chapter focuses on the clinician's decision making and implementation of interventions for OCD. We address indications and contraindications for ERP, when to consider combining ERP with cognitive therapy and medications, various formats for delivering ERP, how to do a functional analysis and select treatment targets, and the influence of comorbidity on treatment decisions. We also review motivational strategies, whether to involve family 31
members in exposure treatment, other aspects of the patient's life that can influence treatment, and finally, how to help patients maintain their own gains. Our recommendations for ERP implementation are based on research findings that we describe briefly in each section to set the stage for clinical decision making.
JUDGING APPROPRIATENESS FOR EXPOSURE AND RESPONSE PREVENTION In general, ERP should at least be considered for all patients with OCD, given the compelling evidence of its efficacy (e.g., Cottraux, Mollard, Bouvard, & Marks, 1993; Fals-Stewart, Marks, & Schafer, 1993; Foa et al, 2005; Lindsay, Crino, & Andrews, 1997; van Balkom et al., 1998). In addition to traditional randomized controlled trials with highly selected participants, ERP has also proved effective in clinical settings (Franklin, Abramowitz, Kozak, Levitt, & Foa, 2000; Warren & Thomas, 2001) and with medicationresistant patients (Kampman, Keijsers, Hoogduin, & Verbraak, 2002; Simpson, Gorfinkle, &. Liebowitz, 1999; Tolin, Maltby, Diefenbach, Hannan, & Worhunsky, 2004). Thus, an expert consensus panel (March, Frances, Carpenter, & Kahn, 1997) wrote that cognitive—behavioral interventions such as ERP are "recommended for every patient with OCD except those who are unwilling to participate" (p. 12). Contraindications for ERP have not been clearly delineated. Some clinicians believe that the exposure element of ERP is too stressful for some patients. Although this issue has not been assessed formally for OCD, a survey of psychologists about treatments for another anxiety disorder, posttraumatic stress disorder (PTSD), revealed that the majority were reluctant to use exposure therapy for severely suicidal or homicidal patients, or for those with a comorbid psychotic or dissociative disorder. In addition, the majority of the sample believed that exposure therapy created a risk in some patients for dissociation, substance abuse, and suicidality (Becker, Zayfert, & Anderson, 2004). In addition (and somewhat contrary to their blanket recommendation), the OCD expert consensus panel (March et al., 1997) did not recommend CBT as a first-line treatment component for patients with comorbid schizophrenia, although they did recommend CBT for patients with all other comorbid disorders, including heart disease. Does exposure-based therapy lead to the exacerbation of anxious symptoms or create an unacceptable risk of adverse events in dissociative, suicidal, homicidal, substance-abusing, or psychotic patients? These issues have not been explored empirically in the treatment of OCD. In a study of exposure-based treatment of PTSD patients, only a minority experienced an initial increase in symptoms; it is interesting to note (and contrary to what some would predict) that patients who experienced symptom exacerbation 32
TOLIN AND STEKETEE
were no more likely than others to drop out of treatment and appeared ultimately to benefit just as much as did those whose symptoms did not increase (Foa, Zoellner, Feeny, Hembree, & Alvarez-Conrad, 2002). In patients with PTSD and comorbid cocaine dependence, many treated with PTSD-directed exposure therapy dropped out of treatment, a response that may not be unique to exposure therapy. However, patients who remained in treatment experienced a decrease in not only their PTSD symptoms but also their cocaine use (Brady, Dansky, Back, Foa, & Carroll, 2001). This result is the opposite of what would be expected if exposure therapy increased the risk of substance use. Thus, there is little reason from an empirical perspective to believe that well-conducted ERP creates a high risk for treatment-emergent adverse events, even in vulnerable individuals, although many clinicians refrain from using these techniques for these reasons. Another factor in making treatment decisions for OCD patients is to consider variables that have been shown to predict poor response to treatment. That is, even if ERP is not deemed risky, this treatment may not be particularly effective for patients with certain features. However, no reliable markers of treatment response have been identified for ERP in OCD patients. Some studies have found that higher initial severity of OCD symptoms was associated with poorer outcomes (de Haan et al., 1997; Keijsers, Hoogduin, & Schaap, 1994), whereas others have not found such an association (Cottraux, Messy, Marks, Mollard, & Bouvard, 1993; Steketee & Shapiro, 1995). Duration of OCD was unrelated to outcome in two studies of ERP (Cottraux, Messy, et al., 1993; Steketee & Shapiro, 1995). Type of OCD may also be related to outcome. Symptoms of compulsive hoarding in particular have been associated with poor response to ERP both with and without medications (Abramowitz, Franklin, Schwartz, & Furr, 2003; Black et al., 1998; Mataix-Cols, Marks, Greist, Kobak, & Baer, 2002; Saxena et al., 2002). Likewise, sexual and religious concerns have also been associated with poor response to ERP (Mataix-Cols et al., 2002), possibly because of poorer insight among patients in these subgroups (Tolin, Abramowitz, Kozak, & Foa, 2001). Research on the effects of comorbid personality disorders is similarly mixed, with some studies finding attenuated treatment response and others finding no attenuation (Fals-Stewart & Lucente, 1993; Steketee, 1990). Some early reports suggested that pretreatment depression predicted poorer outcomes of ERP (Foa, 1979), although a later study indicated that highly and mildly depressed patients responded similarly to treatment (Foa, Kozak, Steketee, & McCarthy, 1992). In a large sample of OCD patients, only severe depression was associated with attenuated outcome of ERP, although even those patients showed significant clinical improvement (Abramowitz, Franklin, Street, Kozak, 6k Foa, 2000). Consistent with these findings, Steketee, Chambless, and Tran (2001) reported that comorbid major depression predicted worse outcomes for patients with OCD or agoraphobia. Lower initial motivation appeared to be associated with poorer outcome of GENERAL ISSUES
33
ERP (de Haan et al., 1997; Keijsers et al, 1994); this may be mediated by reduced follow-through with exposure exercises (de Araujo, Ito, & Marks, 1996; O'Sullivan, Noshirvani, Marks, Monteiro, & Lelliott, 1991). Insight into the irrationality of obsessive fears has been associated with poorer outcome in some studies of ERP (Foa, 1979; Neziroglu, Stevens, & YaryuraTobias, 1999), but not in others (Foa et al., 1983; Hoogduin & Duivenvoorden, 1988). These differences may pertain to the use of different strategies for measuring insight. Ideally, further research on predictors of outcome will lead to the development of treatment algorithms in which patients can be matched a priori to specific treatments. However, the available body of research does not yet support such decisions, with the possible exception of compulsive hoarding, which may require specific interventions tailored to the idiosyncratic nature of hoarding-related symptoms (Hartl & Frost, 1999; Steketee & Frost, 2003).
PHARMACOLOGICAL AUGMENTATION OF EXPOSURE AND RESPONSE PREVENTION The expert consensus panel (March et al., 1997) recommended that for more severe cases of adult OCD, ERP should be combined with serotonin reuptake inhibitor (SRI) medication. The implication, therefore, is that ERP plus SRI should be more efficacious than ERP alone. What do the available data have to say on this topic? To date, four studies have been published that permit a clear test of the efficacy of ERP alone versus ERP + SRI (for a review, see Foa, Franklin, & Moser, 2002). In a comparison (N = 60) of ERP + fluvoxamine (FLV), ERP + pill placebo (PBO), and FLV alone, the combined treatment appeared superior to ERP alone and FLV alone at posttreatment (as measured by ratings of daily rituals). However, at 6-month followup (during which many patients remained on medications), differences among the groups had largely disappeared (Cottraux et al., 1990). At 1-year followup, there remained no significant differences among the groups, although patients who received ERP were less likely to be on medications at that time than were patients who did not receive ERP (Cottraux, Mollard, et al., 1993). A second study (N = 58) of ERP + FLV versus ERP + PBO found a greater number of treatment responders in the ERP + FLV group than in the ERP + PBO group (Hohagen et al., 1998). Outcome was measured using the Yale-Brown Obsessive-Compulsive Scale (Y-BOCS; Goodman et al., 1989). A larger (N = 117) study compared five conditions: cognitive therapy (CT), ERP, CT + FLV, ERP + FLV, and wait list (WL; van Balkom et al., 1998). At posttreatment, there were no significant differences among the four active treatments, and all were superior to WL, although there was a nonsignificant trend toward superiority of combined treatment over ERP or CT monotherapy. There was no follow-up assessment. 34
TOLIN AND STEKETEE
The largest study to date on the topic of ERP alone versus ERP with medications has been published (Foa et al., 2005). The size of this study and the level of detail provided allow a closer look at the comparative effects of ERP monotherapy versus combined therapy. One hundred forty-nine participants were assigned to ERP, clomipramine (CMI), ERP + CMI, or PBO. On learning of their treatment condition, 22% withdrew from the ERP condition, 23% from the CMI condition, and 19% from the PBO condition, but only 6% from the ERP + CMI condition. Although these proportions are not significantly different from one another, they hint at a higher acceptability of combined treatment over either medication or ERP alone. These withdrawals left a sample of 122 patients who entered the study. During treatment, 28% dropped out of ERP, 25% of CMI, and 23% of PBO, but 39% dropped out of ERP + CMI. Again, these proportions were not significantly different from one another, but they suggest that although the combined treatment may be more attractive initially, more patients may discontinue this treatment prematurely. Analysis of outcome using the Y-BOCS after 12 weeks of treatment indicated that patients in all active treatment groups fared better than did those receiving placebo. Patients completing ERP alone experienced a 55% decrease on the Y-BOCS. Patients completing ERP + CMI, CMI, and PBO experienced decreases of 59%, 31%, and 11%, respectively. ERP alone led to significantly greater improvement than did CMI alone, and ERP + CMI was superior to CMI alone. However, the difference between ERP alone and ERP + CMI was not significant. The percentage of treatment-completing patients labeled "treatment responders" using the Clinical Global Impression scale (CGI; Guy, 1976) was 86% for ERP, 79% for ERP + CMI, 48% for CMI, and 10% for PBO. The percentage of patients labeled "excellent treatment responders" using the CGI was 57% for ERP, 47% for ERP + CMI, 19% for CMI, and 0% for PBO. For responder analysis, all active treatments were superior to PBO, and both treatments containing ERP were superior to CMI. However, there were no differences in the number of responders or excellent responders between ERP alone and ERP + CMI. After treatment discontinuation, treatment responders were followed for an additional 12 weeks (Simpson et al., 2004). Relapse rates were 11% for ERP and 14% for ERP + CMI, but 45% for CMI. Thus, patients receiving ERP with or without CMI were less likely to relapse after treatment discontinuation than were patients receiving CMI alone, with no significant difference between ERP and ERP + CMI. In summary, the available data do not clearly answer the question of whether combined therapy is preferable to ERP monotherapy. Those who would argue for ERP monotherapy might note that across studies, although there is a trend for combined therapy to be more effective than ERP alone, this trend does not reach statistical significance and appears to vanish at follow-up. Furthermore, evidence from the largest and most recent trial of GENERAL ISSUES
35
ERP and CMI suggests that combined treatment may be associated with increased risk of dropout without a substantial increase in treatment efficacy. Those who would argue for combination therapy over ERP monotherapy could note several weaknesses in the literature. First, although the available studies do not show a strong advantage for combined therapy, none of the studies speak to the expert consensus panel's recommendation that patients with more severe OCD be given combined therapy right away (March et al., 1997). Second, patients in these studies were generally selected for the absence of certain comorbid conditions such as psychosis, substance abuse, suicidality, or developmental disorders, although other conditions such as depression and personality disorders were usually allowed. Third, studies that randomly assigned patients to treatment conditions might have failed to account for the potentially large impact of patients' preferences for one treatment over the other (TenHave, Coyne, Salzer, & Katz, 2003). Patient preference may influence treatment outcome in several ways, including enrollment and attrition, homework and medication compliance, and expectancy for improvement. Finally, some comorbid conditions, such as severe depression or psychosis, might indicate the use of medications, particularly when comorbidity makes OCD treatment difficult or impossible (e.g., for a depressed patient who is unable to get out of bed and come to therapy reliably). USING COGNITIVE THERAPY TO AUGMENT EXPOSURE AND RESPONSE PREVENTION Currently, the best evidence suggests that ERP is an effective intervention for OCD (Cottraux, Mollard, et al., 1993; Fals-Stewart et al., 1993; Foa et al., 2005; Lindsay et al., 1997; van Balkom et al., 1998). This treatment consists of gradual, prolonged exposure to fear-eliciting stimuli or situations combined with strict abstinence from compulsive behavior. The purpose of these exercises is to allow patients to experience a reduction of their fear response and to recognize that these situations are not excessively dangerous and that their fear will not last forever. Thus, although ERP is a "behavioral" intervention, its mechanism of action may well be cognitive (Foa & Kozak, 1986), and the distinction between behavioral and cognitive therapy may be somewhat arbitrary (Maltby &Tolin, 2003). Should cognitive interventions be used in place of, or as an augmentation of, ERP? During ERP, we routinely assist patients in changing inaccurate beliefs about feared situations, such as pointing out that feared consequences did not occur or that the patient's fear did not persist indefinitely. However, a more formal cognitive intervention teaches patients to identify and correct their dysfunctional beliefs about feared situations (Whittal & O'Neill, 2003; Wilhelm, 2003). In most cases, this has involved either rational-emotive therapy (RET), in which irrational thoughts are identified and targeted via rational debate, or, more recently, cognitive therapy along the lines of Beck et al. (1985), in which Socratic questioning, 36
TOLIN AND STEKETEE
behavioral experiments, and other cognitive strategies are used to challenge the validity of distorted thoughts. The overlap of behavioral experiments with ERP should be clear, and it is possible that the difference is one of emphasis. Certainly, both cognitive therapy and ERP are intended to change both OCD behavior and cognitions, and it is not clear whether the mechanisms of action differ between these two forms of treatment. Although the specific efficacy of cognitive therapy for OCD has not been firmly established, some evidence is promising. In an early study, RET yielded results that did not differ from those of ERP (Emmelkamp, Visser, & Hoekstra, 1988). In comparative efficacy studies of adults with OCD, Beckstyle cognitive therapy produced moderately strong results that also did not differ significantly from those of ERP (Cottraux et al., 2001; van Balkom et al., 1998; van Oppen et al., 1995). In a comparison study using groups to deliver the treatment, cognitive therapy yielded moderate results that were not as strong as those obtained using group ERP (McLean et al., 2001). However, when the same investigators used individual treatment, cognitive therapy was more efficacious and comparable to ERP treatment; both treatments produced high rates of recovery after treatment (67% for cognitive therapy and 59% for ERP; Whittal, Thordarson, &. McLean, 2005). Little empirical attention has been paid to the question of whether the addition of cognitive therapy augments the efficacy of behavioral therapy. An early study of RET (Emmelkamp & Beens, 1991) found that adding this intervention to ERP did not appear to enhance treatment results. In a more recent study (Vogel, Stiles, & Gotestam, 2004), patients were randomly assigned to ERP plus Beckstyle cognitive therapy or ERP plus relaxation training (placebo). Patients receiving ERP plus cognitive therapy were less likely to drop out of treatment than were those receiving ERP plus placebo. However, there was a (nonsignificant) trend for patients receiving ERP plus placebo to show a greater reduction in OCD symptoms, depression, and anxiety. Our preference, on the basis of these data, is to use ERP whenever possible. However, cognitive therapy is an important treatment option when ERP has not produced optimal results or when patients refuse ERP. In an open trial with five adult OCD patients who had failed to respond to pharmacotherapy and ERP, an intensive cognitive therapy program was associated with decreases in self-reported OCD symptoms (Krochmalik, Jones, & Menzies, 2001). Additional trials of cognitive therapy are currently under way (e.g., Wilhelm & Steketee, in press), and findings from these studies should provide a clearer picture of the efficacy of this treatment. OPTIMAL FORMATS FOR EXPOSURE AND RESPONSE PREVENTION There has been little examination of the efficacy of inpatient ERP. Part of the concern with inpatient treatment is that in most hospitals, OCD paGENERAL ISSUES
37
tients are treated in a milieu with non-OCD patients. Treatment is therefore unlikely to be specific to OCD (e.g., formal ERP would be unlikely). However, a small number of inpatient and residential OCD treatment centers have been developed around the United States; these include (but are not limited to) McLean Hospital in Belmont, Massachusetts; the Menninger Clinic in Houston, Texas; and Rogers Memorial Hospital in Oconomowoc, Wisconsin. We are aware of only one systematic attempt to document the effectiveness of these programs (Stewart, Stack, Farrell, Pauls, &. Jenike, 2005). In that study, 486 adult and adolescent OCD patients received an average of 61 days of treatment (ERP and medications), although length of stay ranged as high as 640 days. Among patients with a planned discharge (i.e., excluding those who discontinued the program against medical advice), a 36% decrease in OCD symptoms was noted. An intent-to-treat analysis found a 26% decrease. These decreases seem encouraging, given the likely severity of these patients at admission. However, in the absence of a controlled comparison of outpatient versus inpatient treatment, the higher cost and patient burden of inpatient treatment suggest that OCD should be treated on an outpatient basis when possible. Inpatient treatment may be especially useful for patients who do not respond well to outpatient treatment. ERP delivered by a trained therapist is the treatment of choice for OCD, but this treatment is often difficult to obtain (American Psychiatric Association, 1989; Goisman et al., 1993). ERP is also expensive (in the short term), with a 1995 survey showing an average cost of $4,370 (Turner, Beidel, Spaulding, & Brown, 1995). Although behavior therapy is less expensive over time than longer-term psychotherapy and medications (Otto, Pollack, & Maki, 2000), it still involves considerable expense. Thus, it may be useful in some cases to explore the use of self-administered treatment rather than therapist-administered treatment. There have been comparatively few controlled assessments of self-administered OCD treatment, and many such treatments have actually involved quite a bit of therapist contact (Tolin & Hannan, 2005b). Emmelkamp and Kraanen (1977) found no difference in outcome between therapist-controlled and self-controlled ERP. Their selfcontrolled treatment was directed by the therapist during ten 1 -hour office visits, but the therapist was not physically present during the exposure exercises. Fritzler, Hecker, and Losee (1997) compared partially self-administered treatment patients (Steketee & White's [1990] self-help book plus five sessions of therapist contact to supplement the readings) with wait list patients. Treated patients showed a superior outcome to untreated patients, although only 25% met criteria for clinically significant improvement. Another partially self-driven ERP program is the BT-STEPS program (Baer & Greist, 1997), in which instructions for conducting self-administered ERP therapy are delivered via a computerized telephone administration system. Although BT-STEPS is not purely self-directed (exposure instructions are determined by the computer on the basis of a decision-making algorithm 38
TOLJN AND STEKETEE
using the patient's anxiety ratings), there are no in-person meetings between the patient and a therapist. Open trials found this treatment to be both acceptable to and clinically effective for patients with OCD (Bachofen et al., 1999; Baer & Greist, 1997). In a controlled comparison (Greist et al., 2002), patients were randomly assigned to receive self-administered treatment (BTSTEPS), therapist-administered ERP, or relaxation (placebo treatment). After treatment, 38% of patients in the BT-STEPS group, versus 60% of patients in the therapist-administered treatment group and 14% of placebo patients, were considered responders. The BT-STEPS group showed a 23% reduction on the Yale-Brown Obsessive-Compulsive Scale (Goodman et al., 1989), compared with a 32% reduction for therapist-administered treatment patients and a 7% reduction for placebo patients. It is interesting that when only the treatment-adherent patients were sampled, the therapist-administered and BT-STEPS groups showed similar outcomes, suggesting that the reason for the attenuated results in the BT-STEPS group may have been related to nonadherence to the treatment instructions in that group. Furthermore, despite its moderately positive outcomes, the BT-STEPS program has a very high dropout rate of approximately 50% (Greist et al., 1998), indicating that many patients who entered the trial did not complete it. In an ongoing study of self- versus therapist-administered ERP in one of our clinics, patients were randomly assigned to 15 sessions of therapist-administered ERP or to a commercially available self-help manual (Foa & Wilson, 2001). Preliminary results (Tolin, Hannan, Maltby, Diefenbach, & Worhunsky, 2004) indicate that both groups showed significant improvement in OCD symptoms, with patients receiving treatment from a therapist showing greater improvement (44% Y-BOCS reduction) than did self-help patients (19% Y-BOCS reduction). CGI ratings indicated that 56% of patients receiving therapist-administered treatment were classified as treatment responders, compared with only 15% of those receiving self-administered treatment. Thus, the therapist appears to impart specific benefits over and above the technique of ERP, such as tailored psychoeducation, consultation, support, modeling of exposures, motivation, and accountability for homework compliance (Tolin & Hannan, 2005a). Group therapy may also be a viable method for delivering ERP. Several open trials have demonstrated that ERP in a group setting results in significant decreases in OCD severity (Bouvard, 2002; Krone, Himle, 6k Nesse, 1991; Van Noppen, Pato, Marsland, & Rasmussen, 1998); no difference in efficacy was found between a 7- and 12-week group (Himle et al., 2001). A randomized controlled trial also found group ERP superior to the wait list condition (Volpato Cordioli et al., 2003), and two open trials suggest that group ERP may be an effective intervention for adolescents as well as adults (Himle, Fischer, Van Etten, Janeck, & Hanna, 2003; Thienemann, Martin, Cregger, Thompson, & Dyer-Friedman, 2001). Also encouraging is the fact that even patients with very different OCD symptoms (e.g., washers and GENERAL ISSUES
39
checkers) seem to benefit as much from group therapy as do patients who share the same kind of symptoms (Norton & Whittal, 2004). In an early randomized controlled trial, group and individual ERP were equally effective, although individual therapy appeared to produce faster results (FalsStewart et al., 1993). In a direct comparison, a group-based ERP intervention appeared superior to a group based largely on cognitive restructuring (McLean et al., 2001). In light of a subsequent study of cognitive therapy and ERP delivered individually by the same investigators (Whittal et al., 2005), it appears that cognitive aspects of treatment are more difficult to translate into a group format.
DEFINING AND SELECTING TARGETS FOR TREATMENT The first step in treatment planning is determining whether OCD symptoms or comorbid symptoms need to be addressed first or whether they can be addressed concurrently. Many clinicians adopt a functional approach, targeting the symptoms that appear to be the primary cause of the patient's overall concerns. For example, a patient may present with comorbid OCD and major depression, and during the initial interview, it becomes clear that the onset of depression followed the onset of OCD symptoms. Furthermore, it appears that the depression results primarily from distress and reduced activity subsequent to the OCD. In this case, an argument can be made that the OCD should be treated first, with the expectation that successful OCD treatment will also reduce depressive symptoms. This is, in fact, the finding in most studies; measures of depressed mood decline concurrently with reduction in OCD symptoms (e.g., Abramowitz et al., 2000; Foa et al., 1992). A concern with this approach, however, is that comorbid symptoms, even when secondary to the OCD symptoms, may become severe enough to derail the OCD treatment. So, for example, if the patient were so depressed that he or she could not reliably come to treatment, so fatigued that he or she could not complete exposure exercises, or so suicidal that treatment for OCD put the patient at risk, it would be necessary either to treat the depression first or to treat the OCD and depression concurrently. Thus, what came first is only one aspect of treatment planning; decisions must also take into account the severity of each condition and their interactive effects. Also important in OCD treatment planning is a clear understanding of which symptoms are part of the OCD and which are not. The terms obsession and compulsion can be used quite loosely by the public and clinicians alike, and we have received many referrals for patients with "compulsive" gambling, eating, or sexual behavior or "obsessive" thoughts about wishing to be dead or about prior traumatic experiences. The obsessive thoughts in OCD must be distinguished from the repetitive distressing thoughts characteristic of intense worries in generalized anxiety disorder, ruminations about loss and 40
TOLIN AND STEKETEE
worthlessness in depression, intrusive trauma memories in posttraumatic stress disorder, and many other mental phenomena. Compulsions must be distinguished from impulsive behaviors such as gambling, hair pulling, and shoplifting; stereotyped behaviors such as rocking or head-banging (as might be seen in developmental disorders); or addictive behaviors such as alcohol or drug abuse. Although some authors have characterized all of these as part of an "obsessive-compulsive spectrum" (e.g., Hollander et al., 1996), from a behavioral perspective they are quite different and require different intervention strategies. Part of making this distinction, and also a critical part of early treatment planning, is a functional analysis of behavior. Briefly, a functional analysis consists of identifying the target behavior, external and internal antecedents to the behavior, and immediate and delayed consequences of the behavior. The target behavior in OCD is usually a compulsive or avoidant behavior; mental rituals are also applicable. Antecedents are the factors that seem to trigger the target behavior. These factors may be external to the person (e.g., environmental factors) or internal (e.g., thoughts, feelings, or physiological sensations). In identifying consequences, we emphasize the immediate aftereffects of the behavior that may serve as reinforcers. In most cases, the reinforcement is negative (reduction of an aversive stimulus) rather than positive (introduction of a pleasurable stimulus), as is often the case for impulsive and addictive behaviors. We also examine the delayed and often unintended consequences of the behavior, which may serve as vulnerability factors for later symptoms or may help maintain the person's fears or maladaptive beliefs. Some examples of functional analyses are provided in Table 2.1. Although the OCD symptoms described in Table 2.1 are different, a basic functional pattern is evident in which certain environmental cues lead to intrusive thoughts and feelings of anxiety or tension. The compulsive and avoidant behaviors function not to gratify the person, but rather to reduce these unpleasant feelings. However, these behaviors also block the natural habituation that would normally occur, maintain patients' erroneous beliefs by preventing them from obtaining disconfirming evidence, teach patients that the only way to feel better is to ritualize, and often lead to marked impairment in functioning. Most patients with OCD present with multiple symptom dimensions (Foa et al., 1995). Therefore, the clinician and patient must choose where to begin treatment. We suggest that the primary consideration should be to choose exposure exercises that are feasible and have a high probability of success (i.e., reduced fear). Early treatment successes are one of the more helpful factors that maintain a high level of patient motivation, whereas discouraging experiences early in treatment may lead to nonadherence or dropout. To the extent that moderately easy initial exposures can be identified within multiple symptom domains, we prefer to target the symptoms GENERAL ISSUES
41
CD O C CD D CT CD
co c
«!oi sr-
«i
O CO
!««i 1
CD "a) — O) ^ CO Js tO
"5 o> <jj .= J= 'c g *-
0
"5" o
•H^
s
11 —' 0
-t= .E
._.
8 !® o it"
u>< CO CD
Q
.§ 'S •§ •n *-1
Tl *~
S
-
o cn 'w ^ CO
c ^ "CO
CD
CO CD
'-0 =3
CD CT P CD t CO
E c
—
0 O
0
E 2
CD *O ~
to c ^ CO .zz
f-Q
D5 O) ,—
.c Q. b LL
co "c J=
Behavior
c\i "5 TABLE ponents
o
0) >, LT C ^*
£
D) CD i ~ °-!
-2 "
L?^
LU
C 0c
|s>.
„
lil
C
o|§gr|2
g
^
E ^||
.
^*1 sag P -— Q-
E o O
5^1—